OpenAI's Brockman: Companies Must Act Now on AI Cyber Threats

Brockman laid out a 10-point list of actions companies, which he called "defenders," should take immediately. The list is not detailed in his post, but he stressed that speed is critical: "Time is of the essence, and defenders will need to pursue the steps below at turbo speed."

The immediate implication for corporations: automate security operations. "Over the coming months, every organization will need to begin significantly automating its security program to stay secure," Brockman wrote. He said the security community must urgently create tools and practices that "increase the power of defenders faster than that of attackers as AI continues to advance."

For public companies, the risk is twofold. One, AI-powered attacks could compromise proprietary data, customer information, or critical infrastructure – leading to regulatory penalties, lawsuits, and reputational damage. Two, the cost of defending against such attacks will rise, pressuring margins for firms that rely on manual security reviews.

"The defender's window is open now," Brockman wrote. The window he describes is limited: as AI models improve, the attacker's capabilities will accelerate. Companies that fail to adopt automated defenses soon may find themselves permanently behind.